1990;50:8017-8022. Cancer Res   Sandra H. Bigner, Peter A. Humphrey, Albert J. Wong, et al.   Human Glioma Cell Lines and XenograftsCharacterization of the Epidermal Growth Factor Receptor in

  Updated version

  http://cancerres.aacrjournals.org/content/50/24/8017

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

[CANCER RESEARCH 50, 8017-8022, December 15, 1990]

Characterization of the Epidermal Growth Factor Receptor in Human Glioma CellLines and Xenografts1

Sandra H. Bigner,2 Peter A. Humphrey, Albert J. Wong, Bert Vogelstein, Joachim Mark, Henry S. Friedman, and

Darell D. BignerDepartments of Pathology fS. H. fi., P. A. H.. H. S. F., D. D. B.] and Pediatrics [H. S. F.] and Preuss Laboratory'for Brain Tumor Research [D. D. B.J, Duke UniversityMedical Center, Durham, North Carolina 27710; Fox Chase Oncology Center, Philadelphia, Pennsylvania 19111 [A. J. W.]; Department of Oncology, Johns HopkinsMedical Center, Baltimore, Maryland 21205 [B. y.J; and Department of Pathology, Central Hospital, Skövde,Sweden [J. M.J

ABSTRACT

Both permanent cultured cell lines and athymic mouse xenografts wereestablished from two human glioblastomas. Biopsies from D-245 MGand D-270 MG contained amplified and rearranged epidermal growthfactor receptor (EGFR) genes. Although the gene amplification andrearrangement seen originally was maintained in the xenografts, culturedcell lines established from these biopsies lost the amplified rearrangedgenes in vitro. Analysis of these cell lines and 11 additional permanenthuman glioma cell lines with normal EGFR gene copy number showedfrom 2.7 x Id1 to 4.1 x IO5 high affinity EGFRs/cell by radioreceptor

assay. The RNase A protection assay showed minimal differences in thequantity of EGFR mRNA among the 13 glioma lines, while the D-245MG and D-270 MG xenografts expressed approximately 10-20 times asmuch EGFR mRNA as the corresponding cell lines. Immunoprecipitationof EGFR from these lines, including D-245 MG and D-270 MG, demonstrated only the intact M, 170,000 Da form, while truncated V/,145,000 Da and 100,000 Da EGFR proteins were immunoprecipitatedfrom the D-270 MG and D-245 MG xenografts, respectively. Thesestudies demonstrate that glioma* with amplification of the EGFR geneare capable of establishing in culture but that the amplified rearrangedgenes are not maintained. Possible explanations are that the abnormalgenes are lost during serial passage or that the cells with amplifiedrearranged genes only represent a minor subpopulation of cells, whichare unable to grow in culture. In either case, these observations suggestthat high expression and structural abnormalities of EGFR proteinsgenerated by amplification and rearrangement of the EGFR gene providea growth advantage for gliomas in vivo but not in vitro.

INTRODUCTION

The EGFR3 gene is amplified in 30-40% of malignant human

gliomas (1, 2). Many of these tumors show rearrangement ofthe amplified genes and these abnormal genes are maintainedthrough propagation in athymic mice (3). The products of theamplified rearranged genes are usually smaller than the normalM, 170,000 Da EGFR, due to deletions involving the regionwhich codes for the extracellular, EGF binding region of theprotein (3-5).

Permanent cultured cell lines derived from malignant humangliomas express the EGFR. In contrast to glioma biopsies andxenografts, however, there have been reports of only two lineswith amplified EGFR genes, and the level of amplification hasbeen modest (6, 7). The two abnormal EGFR proteins whichhave been described in vitro are also different from those seenin vivo, in that one was an abnormally large M, 190,000 Daprotein and the other one lacked low affinity receptors and

failed to respond to EGF (7, 8). These findings suggest that,for gliomas, EGFR genes function differently in in vitro and invivo environments.

Here, EGFR genes, transcripts, and proteins are characterized for two human glioma cell lines, D-245 MG and D-270MG, and are compared with xenografts derived from the sameglioma biopsies. Both xenografts contain amplified and rearranged EGFR genes which produce abnormally small EGFRproteins. The EGFR genes are not amplified or rearranged inthe corresponding cell lines and the EGFR proteins which theyexpress are normal in size and function. These observationsdemonstrate that xenografts from gliomas with amplified rearranged EGFR genes are appropriate models for studying theabnormal protein products of these genes, while the relevanceof EGFR gene and protein function in cultured glioma cellsremains to be defined.

MATERIALS AND METHODS

Cell Lines and Xenografts. D-245 MG and D-270 MG cell lines andxenografts were established from glioblastomas from a 70- and a 42-year-old man, respectively. Characterization of the D-245 MG cell lineand both xenografts have been reported previously (3, 9). For comparison, cell lines including the A431 squamous cell carcinoma cell lineobtained from Dr. Christa Stoscheck (Vanderbilt University) and 11established malignant human glioma-derived cell lines were used. Lineswith the prefix "LT were established at the Wallenberg laboratory inUppsala by Ponten and Westermark and lines with the prefix UD" were

established at Duke University Medical Center. All lines except D-336MG have been characterized previously (9-11). D-336 MG was established from a gliosarcoma from a 68-year-old man. The karyotype ofthis original tumor has been published (12).

Cultures are routinely carried in Richter's (13) improved minimal

essential zinc option medium (Grand Island Biological Co., GrandIsland, NY), supplemented with 10% heat-inactivated fetal calf serum,10 mM jV-2-hydroxyethyl piperà/ine jV'-2-ethanesuIfonic acid buffer,and 584 mg glutamine/liter, and are incubated at 37°Cin a 5% CO2

atmosphere. The cell lines are tested every 10th passage for Mycoplasmacontamination by testing the ability of the conditioned medium toconvert deoxyadenosine to adenine (in the presence of an inhibitor ofadenosine deaminase) and thymidine to i Inni ine (14). Upon receipt inthis laboratory, D-343 MGaC12:6, U-373 MG, U-l 18 MG, and U-251MG were contaminated with Mycoplasma. These lines were treatedwith 10 i/H/inl chlortetracycline daily and increasing periods of heatingat 42°Cfor 2 months. Following treatment, they were grown in anti

biotic-free medium for 4 weeks, retested, and found to be free ofMycoplasma. The other cell lines have been consistently negative forMycoplasma contamination.

Chromosomal Analysis. The identity of each cell line and xenograftwas confirmed by karyotype prior to the initiation of these studies.Xenografts were dissociated by incubation for 12-24 h in collagenase(8 mg/ml) in zinc option medium at 37°C.Cultured cells were passaged

Received 6/21/90; accepted 9/14/90.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1This investigation was supported in part by Grants CA-11898 (D. D. B), CA-43460 (B. V.), CA-44640 (H. S. F.). and CA-43722 (S. H. B.) from the NationalCancer Institute, Grant P01-NS-20023 (D. D. B., H. S. F., P. A. H., S. H. B.) and cultured until they achieved exponential growth. Cells in both typesfrom NINCDS, and the Swedish Cancer Society (J. M.). of preparations were arrested in mitosis by treatment with colcemid,

3ÕÕ"T reqUeStSfor;eprin'srSphBUldÕaddr,essed-. . PrP and chromosomal spreads were prepared and G-banded as describedThe abbreviations used are: EGFR, epidermal growth factor receptor; EGF,

epidermal growth factor; DMs, double-minute chromosomes; SDS, sodium do- previously (15).decylsulfate. Tumorigenicity Testing. For D-270 MG, U-343 MGaC12:6, and D-

8017

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

EPIDERMAL GROWTH FACTOR RECEPTOR IN GLIOMAS

336 MG, cells were grown in culture, resuspended in serum-free zincoption medium, and inoculated s.c. into 1-3 athymic mice/cell line at1x10' cells/mouse. Tumorigenicity testing for the other 10 lines has

been reported previously (9, 10).Analysis for Gene Amplification by Hybridization. DNA was purified

from cultured A431 cells and the 13 glioma lines and from frozenxenograft tissue derived from D-245 MG and D-270 MG, using methods detailed previously (16). Five to 7 ¿igof DNA were cleaved withEcoR\, separated by electrophoresis through a 1% agarose gel, blottedonto a nylon membrane, and hybridized with the EGFR probe labeledwith [32P]-dCTP via the random primer method. Washing of filters and

autoradiography were as described (16). The EGFR probe used was the1.6-kilobase EcoRI fragment of pE7, a complementary DNA clone ofEGFR mRNA generously provided by Drs. G. Merlino and I. Pastan(NIH) (17). The signals were removed by boiling, and the filters wererehybridized with a 1.0-kilobase EcoRl/BamHl fragment of pNB containing part of the second exon of the N-myc gene (18).

Ribonuclease Protection. Total RNA was isolated by the acid-guani-dium extraction method described by Chomczynski and Sacchi (19).The integrity of RNA was assessed by denaturing gel electrophoresisand staining with ethidium bromide. 32P-labeled RNA transcripts weregenerated in vitro from the 0.7-kilobase EcoRl fragment of pE7 sub-cloned into the plasmid Bluescript using T3 or T7 RNA polymerase.Ribonuclease protection was performed as described by Winter et al.(20) with the following modifications: hybridizations were performedin a final volume of 10 //I, only RNase A at 12.5 «gml was used, andthe RNase A and proteinase K digestions were performed at roomtemperature for 30 min.

Binding of 125I-EGF to Glioma Cell Lines. Cells were plated in 24-well plates, at 1.0 x 10s cells/well in 1 ml of zinc option mediumcontaining 10% fetal calf serum. Incubation was for 18-24 h at 37°C.

The wells were washed 3 times with 1 ml of binding buffer (20 mM W-2-hydroxyethyl piperazine A"-2-ethanesulfonic acid, 0.1% bovineserum albumin, pH 7.4), and '"I-EGF (Amersham) in binding bufferwas added over a concentration range of 0-25 nM. Each concentrationpoint was assayed in quadruplicate. Nonspecific binding was monitoredby adding a 100-fold molar excess of unlabeled EGF (CollaborativeResearch) to four wells with the highest concentration of '"I-EGF.After a binding reaction of l h at room temperature, unbound '"I-EGF

was aspirated from the plated glioma cells and the wells were washed3 times with binding buffer. The cells were removed from the wells bythe addition of l N NaOH for at least l h at 37°C,followed by transfer

to tubes, which were counted in a Packard Auto Gamma spectrometer.The binding data were analyzed by the Scatchard method on an IBMPC/AT computer using the software program EBDA (equilibriumbinding data analysis), written by G. A. McPherson (21, 22) andobtained from Biomedicai Computer Technology Information Center(Vanderbilt University Medical Center, Nashville, TN 37232).

Immunoprecipitation of EGFR. EGFR from glioma cell lines wassolubilized and immunoprecipitated as described (1), with modifications. Frozen (—70"C)cultured cells (1 x IO7) or 10 mg of xenograft

tissue were homogenized in 1 ml of ice-cold solubilization buffercomposed of 20 mM yV-2-hydroxyethyl piperazine A''-2-ethanesulfonic

acid, pH 7.4, 150 mM NaCl, 1% Triton X-100, 10% glycerol, 4 mMiodoacetate (Aldrich), and aprotinin (Sigma) at 1 mg/ml. After 2 h at4°C,the preparation was centrifuged for 15 min at 4°Cin a Brinkman

tabletop centrifuge at 12,000 x g. The supernatant was used in EGFRimmunoprecipitation. Immunoprecipitation was performed with eithermonoclonal antibody Ab-1 (clone 528; Oncogene Science, Inc.), reactiveagainst the external EGFR domain (23), or polyclonal antibody 2913,reactive with the intracytoplasmic portion of the EGFR (24) andgenerously provided by Dr. I. Pastan, NIH. For each reaction mixture,5 ¿igof monoclonal antibody Ab-1 or 15 ¿ilof undiluted antiserumcontaining polyclonal antibody 2913 were bound to 2 mg of Protein A-Sepharose 4B (Sigma) by incubation in 115 mM sodium phosphatebuffer, pH 7.4, for 30 min at room temperature. Antibody-Protein A-Sepharose complex was washed 3 times with 115 mM sodium phosphatebuffer, pH 7.4. EGFR immunoprecipitation was performed with 500-n\ aliquots of solubilized cells, 500 ^1 of 115 mM sodium phosphatebuffer, pH 7.4, and the antibody-Protein A-Sepharose pellet. After

resuspension, the pellet was incubated overnight at 4°C.Immunopre-

cipitates were washed 3 times with 1 ml of solubilization buffer and thepellet was used for autophosphorylation.

Autophosphorylation of EGFR. Autophosphorylation of the immunoprecipitated EGFR was performed as described (1). The EGFR-antibody-Protein A-Sepharose pellets were incubated with 30 ß\ofsolubilization buffer plus 2 mM MnCI2 and 3 pCi of [7-32P]-ATP (2000-3000 Ci/mmol; New England Nuclear). After 10 min at 4°Con ice, the

reaction was terminated by the addition of 30 ¿ilof 2x Laemmli SDS-polyacrylamide gel electrophoresis sample buffer with 2% 0-mercapto-ethanol. Samples were boiled for 3 min and centrifuged. Supernatantswere used in loading SDS-polyacrylamide gels.

SDS-Polyacrylamide Gel Electrophoresis and Autoradiography. TheSDS discontinuous buffer system of Laemmli (25) with a 7.5% resolvinggel was used. The high molecular weight standard mixture of myosin,rf-galactosidase, phosphorylase /».bovine serum albumin, ovalbumin,and carbonic anhydrase (Sigma) was used to determine apparent molecular weights of the immunoprecipitated EGFR. Gels were stainedwith Coomassie blue R250, destained, and dried under vacuum. Autoradiography was performed by exposing Kodak AR-5 film for varyinglengths of time (1-12 h) at -70°C.

RESULTS

Chromosomal Analysis. Cells of D-245 MG xenograft inpassage 28, D-245 MG cell line in passage 87, and D-270 Mgxenograft in passage 10 had stemlines similar to those reportedpreviously (3, 9). DMs were present in virtually all cells of bothxenografts but could not be found in the D-245 MG cell line.Cells of the D-270 MG cell line were 47,XY,i(6p),t(9ql3q),-+ring,+lDMs (Fig. 1). None of the cell lines or xenograftscontained homogeneously staining regions. Karyotypes pf A431and 6 of the remaining 11 glioma cell lines contained the samestemline karyotypes as had been reported previously (9, 11, 15,26, 27). The five lines which have not had karyotypes publishedpreviously are as follows. D-37 MG: 8=96-97, XX, -X, -X,-10, +11, +19, +20, -22, -22, +del(l)(ql 1), +i(2q), del(7)-(q21), +del(7)(q21), t(9q22q), +t(9q22q), +der(3)(t(3;?)(pl 1;?),der( 13?)t( 13;?) (q32-34;?), der(21 ?)t(21 ;?)(q22;?), +mar, +frags(expressed as deviations from tetraploid). D-336 MG: S=74,XXY, -Y, -2, -4, -5, +7, -9, -9, -10, -10, -11, -12, -13,-14, -14-16, -16, -18, -19, -22, -22, del(3)(q25), del(ll)-

(pll), del(13) (qllq22), der(12)t(9;12)(qll;pll) (expressed asdeviations from tetraploid). U-373 MG: 8=121-122, XXXYY,-4, +5, -10, -11, -13, +14, +14, +14, -15, -18, -18, +19,+20, +21, -22, i(3q), i(3q), del(6)(ql6), del(6)(ql6), del(6)-(q 16), del(6)(q 13q21 ), +del(6)(q 13q21 ), inv(9)(p 11q 13), inv(9)-(pllql3), del(12) (ql5q23), +del(12)(ql5q23), +der(l)t(l;?)-(q31;?), +der(6)t(6;?) (q23;?), +der(16)t(l Ipl6q), +der(16)t-(Ilpl6q), +der(16)t(llpl6q), +der(16)t(l Ipl6q), +DMs (expressed as deviations from pentaploid). U-410 MG: 8=49, XY,+7, +7, -10, del(19)(cen), +der(l)t(lqlOq), +der(19)t(13;19)-(ql4;ql3). U-l 18 MG: 8=96-98, XXY, -Y, -4, -5, +11, +12-16, -18, -19, +20, +20, +22, +i(lq), +del(3)(ql3q33), del(6)-

(ql3q33), +del(6) (ql5), +del(6)(p23qll), +del(6)(p23qll),del(9)(p21), del(9)(p21), del(10)(pl3), del(10)(pl3), +der(3)t(3ql5q)+l-5mars (expressed as deviations from tetraploid).

Tumorigenicity. Cells from U-343 MGaC12:6 grew progressively s.c. in 3 of 3 mice and could be serially transplanted intonew generations of animals. Mice bearing D-270 MG and D-336 MG cells were observed for 1 year, but no tumors grew.Results of tumorigenicity testing for the other 10 lines havebeen reported previously, as follows. U-373 MG, U-l 18 MG,D-54 MG, D-263 MG, D-247 MG, and U-251 MG grew

8018

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

EPIDERMAL GROWTH FACTOR RECEPTOR IN GLIOMAS

progressively when inoculated s.c. into athymic mice (9, 10).D-37 MG, D-245 MG, and D-32/MGaC12:6MG did not. U-410 MG grew initially but regressed (10).

Gene Number and Structure. Hybridization of DNA samplesderived from the 13 glioma cell lines with the EGFR and N-myc probes showed only normal gene copy numbers. In particular, DNA from cell lines D-270 MG and D-245 MG showedno evidence for EGFR gene amplification or rearrangement,while xenografts derived from the same original tumors showedapproximately 10-25-fold amplification of the EGFR gene andretained the same pattern of gene rearrangement as had beendemonstrated in the biopsies from which they had been derived(Fig. 2).

Quantitation of EGFR mRNA. The RNase A protection assay(Table 1, Fig. 3) demonstrated that all of the cultured gliomacell lines varied by no more than 3-fold in EGFR mRNAcontent. In contrast, the D-270 MG and D-245 MG xenograftscontained 8-10 times and >20 times as much EGFR mRNA,respectively, as did the corresponding cultured cell lines andalso contained more mRNA than xenograft D-263 MG, whichdid not have EGFR gene amplification.

Quantitation of Glioma Cell Line EGFR Protein Expressionby Scat chard Analysis of 125I-EGF Binding. Saturation curvesof the direct binding of increasing amounts of 125I-EGF to

cultured glioma cells are shown in Fig. 4. The Scatchard plots

I* U IIx

of these data were usually curvilinear (Fig. 5); the affinityconstants and the number of EGFRs per cell for both the lowand high affinity receptor classes are presented in Table 2. Theglioma cell lines expressed a highly variable number of receptors, from 0.27 x 10" EGFRs/cell (U-251 MG) to 1.6 x IO6EGFRs/cell (D-37 MG). In comparison, the A431 squamous

cell carcinoma line, which is known to express very high levelsof EGFR, had 4.0 x IO6 low affinity EGFRs/cell. The association constant of the EGF-glioma EGFR binding reaction alsovaried over wide range, from 2.5 x 107M~' to 5.6 x IO10M~'.

Apparent Molecular Weight of Glioma EGFR. Fig. 6 exhibitsthe relative sizes of 32P-EGFR immunoprecipitated from cultured glioma cells with monoclonal antibody Ab-1 (528) to theextracellular EGFR domain and with a polyclonal antibodyreactive against the intracytoplasmic portion of EGFR (Ab-2913). The normal-sized M, 170,000 Da EGFR was observedin all six glioma cells lines (Fig. 6, lanes 2-7 and 9-14) and theA431 squamous cell carcinoma line (Fig. 6, lanes 1 and 8); aMr 150,000 Da proteolytic product was present in glioma U-

343 MGaC12:6 (Fig. 6, lanes 3 and 10). The M, 170,000 Dabands in gliomas D-245 MG and D-37 MG are difficult tovisualize at this autoradiographic exposure time (of 0.5 h) butwere readily identified in longer exposures. Five additionalglioma cell lines (U-373 MG, D-45 MG, U-410 MG, U-118MG, and U-251 MG) expressed a normal-sized EGFR with

ft II

il10 11 12

il «t••»i" ",3 X W '5 16 " '8

if

19

«é«f21 22

ring DMsFig. 1. The stemline karyotype of D-270 MG cell line, 47,XY,i(6q),t(9ql3q),+ring,+DMs.

8019

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

EPIDERMAL GROWTH FACTOR RECEPTOR IN GLIOMAS

12345610,000r

- -

-8.0-6.8-5.8

-2.5-2.0-1.8-1.5-1.2

-. — -2.0

Fig. 2. Upper, Southern blot analysis of DNA from A431 cells (lane /),normal human lymphocytes (lane 2), D-270 MG xenograft (lane 3) and cell line(lane 4), and D-245 MG xenograft (lane 5) and cell line (lane 6), cleaved withEcoRi and hybridized with the EGFR probe, demonstrates amplification of thisgene in A431 cells and amplification and rearrangement of this gene in DNAfrom both glioma xenografts. In contrast, only normal copy numbers of unrear-ranged EGFR genes are seen in both glioma cell lines. Lower, the filter shown inthe upper panel rehybridized with an N-myc probe demonstrates similar amountsof DNA in each lane.

Table 1 EGFR mRNA in glioma cell lines and xenografts as determined byRNase A protection

Total cellular RNA was hybridized with an 883-base antisense probe from theEGFR, cleaved with RNase A, and electrophoresed on a 6% acrylamide-urea gel.Relative amounts of EGFR mRNA were determined on autoradiograms bycomparison of glioma cell lines and xenografts with the signal from D-32 ( LMG, which served as the standard and was considered to be 1.

CelllinesD-37

MGU-343MGaC12:6D-247MGD-336MGD-270MGD-54MGU-410MGD-263

MGD-245MGU-251MGD-32

C12MGRelative

mRNAlevel2-33-43-4—

1~1~12-3""I>1-11XenograftsD-270MG

XD-263MG

XD-245MGXRelative

mRNAlevel8-10~1>20

234 7 8 9 10 11 12 13 14 15

Fig. 3. RNase A protection of cell lines and xenografts. Total cellular RNAwas hybridized with an 883-base antisense probe from the EGF receptor, cleavedwith RNase A, and then electrophoresed on a 6% acrylamide-urea gel. The RNAsused were from the cell lines U-410 MG (lane I), U-373 MG (lane 2), U-343MGaC12:6 (lane 3), U-l 18 MG (lane 4), U-251 Mg (lane 5), D-32 C12MG (lane6), D-37 MG (lane 7), D-54 MG (lane 8), D-245 MG (lane 9), D-247 MG (lane10), D-263 MG (lane II), D-270 MG (lane 12), and D-336 MG (lane 13), thexenograft of D-245 MG (lane 14), and the xenograft of D-270 MG (lane 15).

intrinsic kinase activity (data not shown). D-54 MG also expressed the previously described M, 190,000 Da variant (8).The extent of 32P incorporation into the glioma EGFR mole

cule, an indicator of the intrinsic EGFR tyrosine kinase activity,did not correlate with receptor number. Some cell lines withthe higher receptor numbers (A431, U-343 MGaC12:6, and D-247 MG) exhibited a more intense EGFR band than cell lineswith lower receptor numbers (D-336 MG, D-270 MG, and D-245 MG), but other lines, such as D-37 MG, which also had

D-247MG

O 2 4 6 8 10 12 14 16 18 20 22 24 26 28[125I-EGF].„M

Fig. 4. Saturation analysis of'"I-EGF binding to 13 glioma cell lines. Increasing concentrations of 12!I-EGF were added to glioma cells cultured in 24-wellplates. Unbound ' "I-EGF was removed by washing and glioma cells with bound125I-EGF were harvested with l N NaOH. Binding was followed by gamma-

spectrometry.

006 r

004

002

Highaffimi, dos*:=19x10" VT1

i =41«IO9receptori/cell

K„-2 5 «IO7 M"1

Brno. =16*10* receptor!cell

-98 -932 -91

BOUND.LOG M

-895

Fig. 5. Scatchard plot of 125I-EGF binding to the glioma cell line D-37 MG,which expressed the highest level of EGFR. The curvilinear plot suggests twoclasses of EGFR.

Table 2 Estimated affinity and number of EGF receptors on human glioma celllines

Parameters were determined by Scatchard analysis of binding data presentedin Fig. 4.

ReceptorclassHigh

affinityCell

lineA43TD-37

MGU-343MGaCI2:6D-247MGD-336MGD-270MGU-373MGD-54MG*U-410MGD-263MGD-245MGU-118MG*U-251

MGD-32C12MG*K.

(M-1)1.5X

10'1.9x10'3.0x10'1.9x10'7.7x10'1.4x10'2.6x10'2.5

x10'°5.5x10'°5.3x10"3.3

x 10'°Receptors/cell1.3

x10'4.1x10*2.5x10*2.7xIO47.2x10*3.4x10"1.2xIO41.1

xIO44.1xIO31.8xIO40.27

x IO4Low

affinityK.

(M-)1.4

x10'2.5XIO73.7x10"1.8x10'5.9x10'6.7x10s2.1x10"6.7x10"4.8x10'1.7x10'°3.4x10'1.7x10'1.3x10'2.9x 10'Receptors/cell4.0

x10'1.6x10'8.6x10!1.5x10'9.9xIO45.5xIO44.8xIO44.6xIO43.0xIO42.9xIO42.5xIO42.5xIO41.1xIO41.0x IO4

" A431 squamous cell carcinoma line.* Curvilinear Scatchard plots not obtained, so straight line fit was used to

determine the affinity and receptor number for the one class of EGFR.

large numbers of EGFR did not autophosphorylate well (Fig.6, lanes 2 and 9).

Fig. 7 presents a comparative electrophoretic pattern of 32P-

8020

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

EPIDERMAL GROWTH FACTOR RECEPTOR IN GLIOMAS

MW(kDa)

Origin-

205-170-

116-

97-

66-

Ab-1 Ab-2913

t l

1 34567 89 10 11 12 13 14

Fig. 6. Immunoprecipitation of EGFR from solubilized A4.t1 and glioma cellculture homogenates with monoclonal antibody Ab-1 (S28), reactive against theexternal EGFR domain, and with polyclonal antibody 2913 (Ab-2913), reactiveagainst the intracytoplasmic portion of the EGFR. EGFRs in Triton X-100-solubilized homogenates from A431 (lanes 1 and 8) and gliomas D-37 (lanes 2and 9), U-343 MGaCI2:6 (lanes 3 and 10), D-247 MG (lanes 4 and 11), D-336Mg (lanes S and 12), D-270 MG (lanes 6 and 13), and D-245 MG (lanes 7 and14) were immunoprecipitated by the indicated antibody, autophosphorylated with3!P, and analyzed by 7.5% SDS-polyacrylamide gel electrophoresis. Molecularweight standards are indicated on the left margin in kilodaltons (kDa).

MW(kDa)I

origin-

170-

116-97-

66-

Ab-1 Ab-2913C XllC X C XI

1 5 6 8 10

Fig. 7. Immunoprecipitation of EGFR from solubilized cell culture (C) andxenograft (A) homogenates with monoclonal antibody Ah-1 (528), reactiveagainst the external EGFR domain, and with polyclonal antibody 2913 (Ab-2913), reactive against the intracytoplasmic portion of the EGFR. EGFRs inTriton X-100-solubilized homogenates from A431 cell culture (lanes I and 7)and xenograft (lanes 2 and fi), from glioma D-270 MG cell culture (lane 3) andxenograft (lane 4), and from glioma D-245 MG cell culture (lanes 5 and 9) andxenograft (lanes 6 and 10) were immunoprecipitated by the indicated antibody,autophosphorylated with "P, and analyzed by 7.5% SDS-polyacrylamide gel

electrophoresis. Molecular weight standards are indicated on the left margin inkilodaltons (kDa).

EGFR immunoprecipitated from glioma xenograft tissue andglioma cultured cells, which were derived from the same biopsytissue. Monoclonal antibody Ab l (528) immunoprecipitatedthe intact M, 170,000 Da EGFR from both A431 xenografttissue and A431 cell culture (Fig. 7, lanes 1 and 2). The normal-sized receptor was also immunoprecipitated from glioma celllines and D-270 MG (Fig. 7, lane 3) and D-245 MG (Fig. 7,lane 5). In contrast, glioma xenografts D-270 MG and D-245MG expressed a slightly smaller EGFR (D-270 MG) (Fig. 7,lane 4) and an EGFR which is not immunoprecipitated by themonoclonal antibody Ab-1 (528) reactive against the externaldomain (D-245 MG) (Fig. 7, lane 6). D-245 MG EGFR isimmunoprecipitable by the antibody 2913 reactive against the

intracytoplasmic portion of EGFR and is markedly smaller(with a doublet at M, 80,000 Da and 100,000 Da) (Fig. 7, lane10) than the normal-sized M, 170,000 Da EGFR that is expressed by D-245 MG cell culture (Fig. 7, lane 9). The normal-sized M, 170,000 Da D-245 MG cell culture EGFR band inFig. 7, lane 9, was more easily visualized at longer autoradi-ographic exposures and could be seen in Fig. 7, lane 5.

DISCUSSION

Permanent cultured cell lines derived from human neoplasmsprovide useful in vitro models for defining a variety of biologicalproperties of these neoplastic cells. Gene amplification, inparticular, has been studied in experimental settings in vitro,since this property can be maintained only under conditions ofselective pressure.

The human tumor type which is best characterized for geneamplification both in vivo and in vitro is human neuroblastoma.While approximately 38% of resected neuroblastoma containamplification of the N-myc gene, amplification of this genecharacterizes virtually all permanent cultured neuroblastomacell lines, suggesting that it is necessary for these cells toestablish in culture (18, 28). For malignant gliomas, in contrast,30-40% of biopsies contain amplification of the EGFR gene(1, 2). Among the more than 50 glioma cell lines which havebeen analyzed to date, however, there are only two examples ofEGFR gene amplification, and both lines show only a modestincrease in gene copy number (6, 7). There are several theoretical explanations for this observation. The demonstration thatD-245 MG and D-270 MG, two gliomas with amplification ofthe EGFR gene, established in culture excludes the possibilitythat gliomas with amplification of this gene are not capable ofestablishing in culture and negates the possibility that gliomaswith EGFR gene amplification have not been tested in culturefor their ability to establish. For these two tumors, amplificationand rearrangement of the EGFR gene, which was demonstratedin initial biopsies and xenografts, was lost as these tumorsestablished in vitro. This observation suggests that the selectiveconditions in vivo which favor amplification of this gene arenot present in culture. The inability of the D-245 MG and D-270 MG cell lines to grow when injected into athymic micesupports this concept. Furthermore, studies on NIH 3T3 cellsshow that transfection and overexpression of the EGFR canendow this cell line with the ability to grow in athymic mice(29).

The cultured lines D-270 MG and D-245 MG contain thenumber of EGFR gene copies appropriate for their copy numberof chromosome 7, fail to show any rearrangement of this gene,and express appropriate amounts of structurally normal EGFRmRNA and protein for the number of gene copies which theycontain. In this respect, they resemble the other glioma celllines described here, which include three lines for which theinitial biopsies were available and which did not contain EGFRgene amplification in the original human tumors.

Both D-245 MG and D-270 MG biopsies and xenograftscontained rearranged as well as amplified EGFR genes, andthese tumors bound EGF with a lesser capacity than tumorswith amplified, structurally normal EGFR genes (3). This observation suggests that these abnormal proteins may be capableof functioning in the absence of EGF and, therefore, cellsexpressing these abnormal EGFRs may possess a growth advantage over cells expressing only the normal EGFR in thepresence of low levels of EGF. It is not possible to determinethe amounts of EGF accessible to glioma cells //; vivo and inxenografts, and no exogenous EGF was added to the culture

8021

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

EPIDERMAL GROWTH FACTOR RECEPTOR IN GLIOMAS

medium used to generate and propagate these cell lines. Nevertheless, since transforming growth factor a mRNA is frequentlyexpressed by glioma cell lines in culture (11), the possibilityremains that higher levels of EGFR ligands may be present Invitro than in in vivo and thus may remove the selective advantageimparted by the amplified rearranged EGFR genes.

Many glioma cell lines express platelet-derived growth factorand/or its receptor (11). Platelet-derived growth factor as wellas other growth factors and hormones are present in fetal calfserum, raising the alternative possibility that differences in thelevels of these substances between the in vitro and in vivosettings may be the selective pressure producing the loss ormaintenance of EGFR gene amplification and rearrangementin these settings, respectively. Attempts to establish culturedcell lines from the D-270 MG and D-245 MG xenografts indefined media containing various growth factors may allowmaintenance of the amplified and rearranged genes and maypermit precise identification of the factors which interact withthe EGFR in gliomas.

The observation that the amplified and rearranged EGFRgenes were lost as D-245 MG and D-270 MG established inculture but were maintained in xenografts is in contrast toexamples of squamous cell carcinomas such as A431 cells,where amplification of the EGFR gene appears to provide agrowth advantage in vitro (30). One explanation for this difference is that the EGFR genes and proteins may function differently in glioma cells than in squamous cell carcinoma cells.Alternatively, the difference may be due to the nature of therearrangements of the amplified EGFR genes in these twogliomas.

The high incidence of amplification and rearrangement ofthe EGFR gene in human gliomas is an interesting and undoubtedly biologically significant property of these tumors.Further analysis of the glioma cell lines and xenografts described here may provide insight into the way in which theseamplified genes function in the in vivo and in vitro settings.

REFERENCES

1. Libermann, T. A., Razón,N., Bartal, A. D., Yarden, Y., Schlessinger, J., andSoreq, H. Expression of epidermal growth factor receptors in human braintumors. Cancer Res., 44: 753-760, 1984.

2. Wong, A. J., Bigner, S. H., Bigner, D. D., Kinzler, K. W., Hamilton, S. R.,and Vogelstein, B. Increased expression of the epidermal growth factorreceptor gene in malignant gliomas is invariably associated with gene amplification. Proc. Nati. Acad. Sci. USA, 84: 6899-6903, 1987.

3. Humphrey, P. A., Wong, A. J., Vogeistein, B., Friedman, H. S., Werner, M.II., Bigner, D. D., and Bigner, S. H. Amplification and expression of theepidermal growth factor receptor gene in human glioma xenografts. CancerRes., ¥«.'2231-2238,1988.

4. Yamazaki, H., Fukui, Y., Ueyama, Y., Tamaoki, N., Kawamoto, T., Tani-guchi, S., and Shibuya, M. Amplification of the structurally and functionallyaltered epidermal growth factor receptor gene (c-erft-B) in human braintumors. Mol. Cell. Biol., f!: 1816-1820, 1988.

5. Maiden, L. T., Novak, U., Kaye, A. H., and Burgess, A. W. Selectiveamplification of the cytoplasmic domain of the epidermal growth factorreceptor gene in glioblastoma multiforme. Cancer Res., 48:2711 -2714,1988.

6. Filmus, J., Pollak, N., Cairncross, J. G., and Buick, R. N. Amplified,overexpressed and rearranged epidermal growth factor receptor gene in ahuman astrocytoma cell line. Biochem. Biophys. Res. Commun., 131: 207-215, 1985.

7. Wells, A., Bishop, J. M., and Helmeste, D. Amplified gene for the epidermalgrowth factor receptor in a human glioblastoma cell line encodes an enzy-matically inactive protein. Mol. Cell. Biol., 8:4561-4565, 1988.

8. Steck, P. A., Lee, P., Hung, M-C, and Yung, W. K. A. Expression of analtered epidermal growth factor receptor in human glioblastoma cells. CancerRes., 48: 5433-5439, 1988.

9. Bigner, S. H., Friedman, H. S., Biege!, J. A., Wikstrand, C. J., Mark, J.,Gebhardt, R., Eng, L. F., and Bigner, D. D. Specific chromosomal abnormalities characterize four established cell lines derived from malignant human gliomas. Acta Neuropathol., 72:86-97, 1986.

10. Bigner, D. D., Bigner, S. H., Pontén,J., Westermark, B., Mahaley, M. S.,Jr., Ruoslahti, E., Herschman, H., Eng, L. F., and Wikstrand, C. J. Heterogeneity of genotypic and phenotypic characteristics of fifteen permanent celllines derived from human gliomas. J. Neuropathol. Exp. Neurol., 15: 201-229, 1981.

11. Nister, M., Libermann, T. A., Betsholtz, C., Pettersson, M., Claesson-Welsh,L., Heldin, C-H., Schlessinger, J., and Westermark, B. Expression of messenger RNAs for platelet-derived growth factor and transforming growthfactor a and their receptors in human malignant glioma cell lines. CancerRes., «.-3910-3918, 1988.

12. Bigner, S. H., Mark, J., Burger, P. C., Mahaley, M. S., Jr., Bullard, D. E.,Muhlbair, L. H., and Bigner, D. D. Specific chromosomal abnormalitiescharacterize malignant human gliomas. Cancer Res., 48:405-411, 1988.

13. Richter, A. Improved medium for the cultivation of differentiated linear cells.In Vitro, 6: 220-221, 1970.

14. Kurtzberg, J., and Hershfield, M. S. Determinants of deoxyadenosine toxicityin hybrids between human T and B lymphoblasts: a model for the development of drug resistance in T cell acute lymphoblastic leukemia. Cancer Res.,45: 1579-1587, 1985.

15. Bigner, S. H., Mark, J., and Bigner, D. D. Chromosomal composition offour permanent cultured cell lines derived from human gliomas. CancerGenet. Cytogenet., 10: 335-349, 1983.

16. Goelz, S. E., Hamilton, S. R., and Vogelstein, B. Purification of DNA fromformaldehyde-fixed and paraffin-embedded human tissue. Biochem. Biophys.Res. Commun., 130: 118-126, 1985.

17. Xu, Y-H., Ishii, S., Clark, J. L., Sullivan, M., Wilson, R. K., Ma, D. P., Roe,B. A., Merlino, G. F., and Pastan, I. Human epidermal growth factor receptorcDNA is homologous to a variety of RNAs overproduced in A431 carcinomacells. Nature (Lond.), 309:806-810, 1984.

18. Schwab, M., Alitalo, K., Klempnauer, K-H., Varmus, H. E., and Bishop, J.M. Amplified DNA with limited homology to myc cellular oncogene isshared by human neuroblastoma cell lines and a neuroblastoma tumour.Nature (Lond.), 305: 245-248, 1983.

19. Chomczynski, P., and Sacchi, N. Single-step method of RNA isolation byacid guanidinium thiocyanate-phenol-chloroform extraction. Anal.Biochem., 162: 156-159, 1987.

20. Winter, E., Vaniamolo, F., Almoguera, C., and Perucho, M. A method todetect and characterize point mutations in transcribed genes: amplificationand overexpression of the mutant c-Ki-ros alíelein human tumor cells. Proc.Nati. Acad. Sci. USA, 82:7575-7579, 1985.

21. McPherson, G. A. A practical computer based approach to the analysis ofradioligand experiments. Comput. Prog. Biomed., 17:107-114, 1983.

22. Munson, P. J., and Robard, D. LIGAND: a versatile computerized approachfor the characterization of ligand binding systems. Anal. Biochem., 707:220-239, 1980.

23. Kawamoto, T., Sato, T. D., Le A. D., Polikoff, J., Sato, G. H., and Mendelsohn, J. Growth stimulation of A431 cells by epidermal growth factor:identification of high-affinity receptors for epidermal growth factor by ananti-receptor monoclonal antibody. Proc. Nati. Acad. Sci. USA, 80: 1337-1341, 1983.

24. Beguinot, L., Werth, D., Ito, S., Richer!, N., Willingham, M. C., and Pastan,I. Functional studies on the EGF receptor with an antibody that recognizesthe miraceli ular portion of the receptor. J. Biol. Chem., 261: 1801-1807,1986.

25. Laemmli, U. K. Cleavage of structural proteins during the assembly of thehead of bacteriophage TV Nature (Lond.), 227:680-685, 1970.

26. Mark, J., Westermark, B., Pontén,J., and Hugosson, R. Banding patternsin human glioma lines. Hereditas, 87: 243-260, 1977.

27. Wikstrand, C. J., Bigner, S. H., and Bigner, D. D. Demonstration of complexantigenic heterogeneity in a human glioma cell line and eight derived clonesby specific monoclonal antibodies. Cancer Res., 43: 3327-3334, 1983.

28. Brodeur, G. M., Seeger, R. C., Schwab, M., Varmus, H. E., and Bishop, J.M. Amplifications of N-myc in untreated human neuroblastomas correlateswith advanced disease stage. Science (Washington, DC), 224: 1121-1124,1984.

29. DiFiore, P. P., Pierce, J. H., Fleming, T. P., Hazan, R., Ullrich, A., King,C. R., Schlessinger, J., and Aaronson, S. A. Overexpression of the humanEGF receptor confers an EGF-dependent transformed phenotype to NIH3T3 cells. Cell, 51: 1063-1070, 1987.

30. Santon, J. B., Cronin, M. T., MacLeod, C. L., Mendelsohn, J., Masui, H.,and Gill, G. N. Effects of epidermal growth factor receptor concentration ontumorigenicity of A431 cells in nude mice. Cancer Res., 46: 4701-4705,1986.

8022

Research. on January 3, 2014. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from